The Economic Theory of'Managerial' Capitalism
What set Marris off on the line of research that was to make his name is unclear—it was not a topic on which others in Cambridge were working. The likely explanation is that an interest in the behaviour of firms acquired in his study of shift work led him to read and reflect on three important books about the firm that were published in the late 1950s. His thinking progressed during a 1960-1961 sabbatical at Austin, Texas, and Berkeley, including seminars at other US universities, where he forged enduring links with many leading American economists.
Marris’s focus was on the separation of ownership from management in joint-stock companies (or ‘corporations’) and on the weakness of control by shareholders, which gave managers considerable discretion. This feature of corporations, which produced a large share of gross domestic product, had been highlighted by Berle and Means (1932) and again by Burnham (1941) and Gordon (1945), but with no impact on the theory of the firm in economics texts, which continued to assume that markets were usually supplied by a multiplicity of small owner-managed firms. Nor had economic theory come to terms with the challenge of Sraffa (1926), underlined by Kaldor (1934), to Marshall’s view that decreasing returns to scale caused there to be an optimal size of the firm.
Three independent contributions to a new theory then arrived simultaneously. Baumol (1959), influenced by his experience as a consultant, produced a formal model of the firm in which the personal benefits to managers of running larger firms caused them to maximise sales revenue subject to a minimum profit constraint. Penrose (1959) brilliantly recharacterised the firm, not as the supplier of some specific good but as a team of businesspeople able to produce many goods and enter new markets by merger or diversification. Abandoning the concept of optimal size, her theory focused on the growth of the firm, and on limits to the speed of its growth—though only informally, as Marris (1961) noted in a complimentary review.[1] In an analysis of competition as a dynamic process, Downie (1958: 63-67) also sketched a model of the firm in which managers want to maximise growth subject to various dynamic constraints.
Building on this earlier work, and taking returns to scale to be constant, Marris set out a formal model of the growth of a firm run by managers who want it to grow as fast as possible. He also explored the implications of his model for competition, economy-wide growth, and welfare. Almost all his publications in this field were in the period 1963-1980, but in 1998, he produced Managerial Capitalism in Retrospect, a revised and abridged edition of his famous 1964 book, The Economic Theory of‘Managerial’ Capitalism.[2] The 1998 version, which also contains Marris’s final assessment of work in this field, is the best point of entry into this aspect of his writings.
The first version of his model (Marris 1963) appeared as a lead article in the Quarterly Journal of Economics (QJE), with a star-studded list of acknowledgements to economists in Cambridge and elsewhere. Managers are argued not only to want to make their firms grow—to increase their salaries, status, and power—but also to be concerned about the risk of being dismissed after a successful takeover raid. Three dynamic constraints on the firm’s growth are identified and modelled. One is the cost of diversification into new markets, which causes the firm’s average profit margin on sales to decline with faster growth. A second is the internal organisational costs of growth (the ‘Penrose effect’), which cause the firm’s capital-output ratio to rise with its growth rate.
The third constraint is the need to generate finance for investment in new capacity. As borrowing is limited by risk to some ratio of capital and new issues of shares are rare (and so assumed to be zero), the growth rate of the firm is linked positively to its profit rate by the retention ratio—the share of profits not distributed as dividends—to which an upper limit is set by the fear of takeover. The three constraints, expressed as equations, are ingeniously stitched together in a diagram in which the growth rate of the firm’s assets is traced as an inverse U-shaped function of its diversification rate, of which the managerial desire for growth causes the apex to be chosen.
In a series of publications over the next two decades, this model evolved in various ways. It is convenient, however, to jump to its final version in Chapter 6 of the 1998 book (which is almost identical to the form it had reached by 1980). In the diagram that sums up this version (Figure 40.1), the firm’s growth rate is on the horizontal axis, while the vertical axis shows its valuation ratio—the stock market value of its shares divided by the book value of its capital stock. Faster growth at first increases but then decreases the valuation ratio. Shareholders would like to be at the peak of this curve, but managers would prefer a higher growth rate. How far down the slope they choose to go is constrained by their fear of takeover, which sets a lower limit (assumed in this figure to be unity) on the valuation ratio and thus fixes the growth rate.
It is of interest to consider what stayed the same between the first and last versions of the model, what changed, and why. One obvious enduring feature
Figure 40.1 Determination of firm's growth rate is the assumption, explained most fully in Chapter 2 of the 1964 book, that managers want to maximise the growth rates of their firms without running too much of a risk of dismissal as the result of a takeover. Another enduring feature is the upper limit on the gearing ratio—the ratio of debt to capital—which restricts growth by borrowing (although the utility of managers is assumed to depend on the scale of gross assets, not of equity capital).
All versions include a decreasing relationship between growth and profitability as a result of the rising cost of diversification, which is discussed and cleverly modelled in Chapter 4 of both editions of the book. But its rationale was broadened to include the costs of all the ways of expanding demand for the firm’s products—research, advertising, acquisitions—while keeping its prices competitive. Moreover, the distinction between the effects of growth on the profit margin and on the capital-output ratio (originally suggested to Marris by Machlup) disappeared, and the ‘demand’ relationship became simply that between the growth rate and the profit rate on capital. Marris also came to believe that this relationship was decreasing only beyond a certain point, up to which the dynamics of growth increase profitability. In addition, he came to emphasise more clearly that his was a profit-driven model— growth-maximising firms were efficient in every sense other than growing faster than their shareholders would have wished.
What changed more radically was the ‘finance’ side of the model. An obvious new arrival, which made its entry in the 1964 book, is the valuation ratio (later reinvented by Tobin as his celebrated q). This variable, which required a formal treatment of the determination of share prices, was introduced to make the interests of shareholders—and their conflict of interest with managers— clearer than in the 1963 article. It also permitted a fuller analysis of the risks of takeover, of which the 1964 book contained the first formal theory and laid the basis for much later empirical work.[3]
A notable disappearance was that of the financial variables—the retention ratio and the rate of issuance of new shares. Even before the QJE 1963 article, Modigliani and Miller had argued that shareholders should be indifferent to whether investments were financed by retaining profits or issuing new shares—all that mattered to them was the profitability of the investments. However, the assumptions required for this to be true, including complete certainty, seemed highly unrealistic to Marris, so in his 1964 book the valuation ratio depends both on profitability and on the financial variables.
Shortly after the book appeared, however, Marris’s Cambridge and King’s colleague, Richard Kahn, worked on the algebra of the valuation ratio and generated the same result as Modigliani and Miller, which caused Marris to capitulate (wrongly, in the present author’s view).[4] In later versions of his model, the retention ratio and the new issue rate play no substantive role.[5] Kahn’s formula, however, had the drawback that as the growth rate of a firm approached the market discount rate its valuation became infinite. Marris eventually solved this problem by making the firm’s market discount rate increase with its growth rate, building on the analysis of share pricing under uncertainty in Lintner (1971).
In several publications, Marris explores the implications of his theory of the firm for competition and concentration in product markets—recon- necting with the origins of his theory in Downie’s innovative analysis of the competitive process. He saw firms as competing at two levels (Marris 1998a: xix). At a low level, there is competition among the divisions of corporations (or quasi-firms) in mature markets—which governs the ‘basic’ profits of firms, before the absorption of part of these profits by spending on diversification and other demand-expanding activities. In Marris and Wood (1971: Chapter 9), he outlines a more general theory of oligopoly, based on the threat of entry, although most of his work on imperfect competition was done much later (Marris 1991).
At the higher level, to which Marris devotes far more attention, there is competition among whole firms, though of a sort which tends to generate increasing concentration, a feature of capitalist economies that in his view is not explained by the neoclassical theory of the firm. In seeking to understand the concentration process, he repeatedly studied data on the persistence of firm profitability and growth and on the correlations among profit rates, growth rates, and firm size.[6]
Marris’s fullest contribution in this field is his de Vries Lectures (1979), summarised and extended in Marris and Mueller (1980: 47-50), in which he combines his model of the firm with the stochastic theories of concentration pioneered by Gibrat and Steindl. More specifically, he shows that a stochastic version of his model generates a precise and predictable path of increasing concentration through internal growth and merger (simulated with a size-related takeover probability matrix). The trajectory of this path depends crucially on whether firm efficiency (in generating profits and using them for diversification) tends to increase or decrease with firm size.
The macroeconomic implications of his theory of the firm are also a recurrent theme of Marris’s work.[7] He believed strongly that the activities of managerial firms, which stimulated the expansion both of demand and of supply, had accelerated the growth of capitalist countries in the twentieth century. His ideas were skilfully elaborated by Odagiri (1981). One of their key policy implications is paradoxical: growth would be slowed by more competition, both in product markets (reducing the basic profits available to spend on diversification) and in the stock market (increasing the threat of takeover). Marris (1998a: xix, 164-165) emphasised the association of fast growth in post-war France, Germany, and Japan with the virtual absence of involuntary takeovers.
A side effect of this line of argument was to lead Marris into disagreement with his Cambridge colleagues about the relationship between the rate of growth and the rate of profit. The position of Kaldor, Pasinetti, and others, including the present author, was that the rate of profit is determined by the rate of growth, which could be taken as exogenous. The original Kaldor- Pasinetti linkage from growth to profits was through differential saving rates, but Kaldor (1966), without attribution to Marris, set out a new model using the retention ratio, share issues, and the valuation ratio.
Marris’s views were entirely different. In Marris (1972), as in Marris (1964a: 309-310) and Marris (1991: 203-210), he maintains that the rate of profit is exogenously determined by ‘the degree of oligopoly’ and is a determinant of the growth rate rather than being determined by it. The behaviour of managers also affects both growth and profits (since reported profits are net of diversification spending and thus understate the degree of market power: Marris and Mueller 1980: 56). Marris saw no basis for assuming differential saving rates and argued that Kaldor could not treat the retention ratio as exogenous, since it was set by managers to generate enough finance for their chosen growth rates. He also repeatedly criticised all the Cambridge macro theories of distribution for their failure to provide adequate microfoundations.[8]
Although he believed that managerial capitalism resulted in faster growth, Marris was far more cautious about its broader normative implications. He was doubtful about the effects on economic welfare of diversification and other means of demand creation, which are central to his theory of the firm. Useful new products benefit consumers, but other ways of creating demand may be wasteful or harmful: ‘The possibility thus arises that the total cost of investments in nonprice competition exceed the benefits’ (ibid.: 57). Marris had even more doubts about the social and political merits of managerial capitalism, ending a chapter on the institutional needs of a better society with the words, ‘Increasingly, I think, the days of the stockholder corporation must be numbered’ (Marris 1974: 399).
In his 1998 book, looking back over 35 years, Marris recognises that economics had moved on. Building on the insights of Coase, of which he had initially been unaware, the existence of large firms can now be theorised in terms of their ability to reduce the high costs of transactions among small firms—though Marris found this explanation more plausible for vertically than for horizontally integrated firms (Marris 1998a: xv). The separation of ownership from control can now also be theorised as just one example of the difficulties that arise when principals have to act through agents.
Marris portrays his work as contributing to one of three strands—‘growth- orientated’—of managerial theories of the firm. The other strands are ‘discretionary’ theories, exemplified by Baumol and Oliver Williamson, in which managers use their freedom of manoeuvre to feather their own nests, and ‘bureaucratic’ theories, exemplified by Monsen and Downs, in which managerial control just makes large firms inefficient. Within the growth-orientated strand, started by Penrose, Marris stresses two contributions by Mueller, of which one was to develop the theory of takeovers from the perspective of the acquiring (rather than, as in Marris’s model, the acquired) firm.
Mueller’s other contribution was a life cycle interpretation of firm growth. In a firm that is ‘young’, in the sense of having found or created an unusually profitable set of new markets, there may be no trade-off between faster growth and a higher valuation ratio—over any relevant range, the two variables move together. Only when a firm ‘ages’, after saturating its initial markets, does a conflict arise between the desire of managers for faster growth and the desire of shareholders for higher valuation.
Marris recognised that there were prominent real examples of ‘young’ and ‘dynamic neoclassical’ firms, usually owner controlled, in which there is no trade-off between growth and valuation and sometimes stellar performance. He also acknowledged the relevance of the discretionary and bureaucratic theories—there are sleepy managerial firms as well as the dynamic growth-seeking ones on which he focused. He continued to believe, however, that dynamic managerial firms had been the most common type of business corporation during the twentieth century, and he argued that empirical studies strongly support managerial theories (Marris 1998a: 117-119; see also Hughes 1987).
In his 1998 book, Marris also recognises that the world had changed in three notable respects. One is that the share of managerial remuneration in the form of stock options and bonuses has increased, partly with the aim of aligning the interests of managers more closely with those of shareholders. Marris, however, was sceptical about stock options: as he put it, noting the greatly increased pay of top managers, ‘once there is separation of ownership from management, the search for a perfect control contract between shareholders and managers is...liable to be perverted’ (Marris 1998a: 40).
A change that he acknowledged may have made more of a difference was the greater ease of takeovers - more frequent mergers, more takeover raiders and the use of junk bonds as a financing vehicle - tending to push the choices of managers closer to the peaks of their valuation-growth frontiers. But he did not accept that this change had turned most corporations into valuation maximisers: the majority of raiders were other managerial firms, seeking to get bigger, a hypothesis strongly supported by evidence that, in most cases, takeovers make the shareholders of the acquiring firm worse off.
The third change was the information technology (IT) revolution. Marris noted that many IT firms were young or of the dynamic neoclassical type. Many IT firms also had little physical capital and incentivised their employees with allocations of shares rather than high salaries. In addition, Marris noted that IT itself had altered key determinants of the advantages and disadvantages of firms being large rather than small, although he was uncertain about the direction of the net effect. In so far as he believed that firms in the twenty-first century might behave differently from those in the twentieth century, it is IT that he saw as the most likely cause.
At the time of writing this chapter, it is just over half a century since Marris’s landmark book appeared. His work precipitated a major academic debate that drew in Arrow, Solow, and other leading mainstream economists and became known to the public. The book also continues to be cited—about 75 citations per year since 2000.[9] Few of the recent citations, however, are in mainstream economics publications, tending to confirm Marris’s own 1998 assessment of the long-term impact of his work.[10] Though ‘bloody but unbowed’ (ibid.: xvi), he felt that his effort to divert economic theorising from the sizes of firms to their rates of growth had been ‘without great influence’ (ibid.: 113). It is to be hoped that his original and important work in this area will be rediscovered.
- [1] Marris’s review, probably written in late 1960 during his US sabbatical, stresses the need for a formaltheory of the growth of the firm, but gives no hint that he was working on one.
- [2] Recognising that the 1964 book had been work in progress, the abridgement cut out about half of theoriginal text and removed most of the original algebra. Its lively style, however, was maintained.
- [3] In most versions of the model, the takeover risk constraint is expressed as some minimum value of thevaluation ratio, but in a few versions the managerial utility function has a continuous trade-off betweenthe growth rate and the valuation ratio.
- [4] Kahn’s analysis of the valuation ratio, published later in Kahn (1972), is based on assumptions just asunrealistic as those of Modigliani and Miller. Wood (1975) argues against these assumptions and putsforward a theory of profits (at both firm and economy-wide levels) in which the financial variables play akey role.
- [5] Though it was sometimes convenient for the purposes of exposition to assume no new issues and use theretention ratio to link the growth rate to the profit rate (e.g. Marris and Mueller 1980: 47).
- [6] For example, Marris (1964a: 277-288), Marris and Wood (1971: Appendix B), and Marris (1972: 330).
- [7] Marris (1964a: Chapter 8), Marris (1972: Appendices I and II), Marris and Mueller (1980: 45-46),and Marris (1998a: xix and Chapter 8).
- [8] Agreeing with Marris about the lack of microfoundations, the present author tried to provide them inWood (1975), but in the opinion of Marris (1991: 205) ‘misguidedly’. Marris’s criticism stemmed in partfrom disagreement about financial variables (see fn. 7 above), and in part from the absence from Wood(ibid.) of any formal model of price determination, to which the present author’s response is that histheory of profits sought to explain reported profits, not basic profits.
- [9] These citation data are from Google Scholar and refer to the 1964 book rather than the 1998 revision,which has only 51 citations in total.
- [10] In none of the volumes of the North-Holland Handbook of Industrial Organisation does ‘managerialcapitalism’ appear in the subject index. In the 1989 volume I, Chapter 2 refers to ‘the older literature onmanagerial theories of the firm’, but without mentioning Marris.
